Performance Comparison of Video Protocols using Dual-Stack and Tunnelling Mechanisms Hira Sathu 1,1, Mohib A. Shah 1, Kathiravelu Ganeshan1 1
[email protected],
[email protected],
[email protected] 1
Unitec Institute of Technology, Auckland, New Zealand
Abstract. This paper investigates the performance of Video Protocols over IPv6 and IP transition mechanisms. It mainly focuses on the impact caused by IP transition mechanisms on video packets and compares this with pure IPv6 based networks. The video protocols selected in this experiment were MPEG-1, MPEG-2, MPEGP-4, MKV and FLV. In this experiment a Dual-Stack and two tunnelling mechanisms were established and the impact of these mechanisms on five video protocols was measured. The parameters measured were actualthroughput over a pure IPv6 network, impacted-throughput (due IP transition mechanisms) and CPU utilization. The results indicate that video packet with large size had been impacted more than packets with small size using these IP transition mechanisms. Dual-Stack mechanism performed much better than two tunnelling mechanisms (IPv6to4 & IPv6in4) tested. IPv6in4 tunnelling mechanism had more impact than IPv6to4 tunnelling mechanism over all the video protocols tested with IPv6to4 marginally close for all protocols tested. Performance comparison between video protocols shows that FLV protocol was least impacted while MPEG-2 was highly impacted by the tunnelling mechanisms. Further detail is covered in this paper including specification for actual-throughput, impacted-throughput and CPU utilization. Keywords: Video protocols, performance evaluation, actual-throughput, Impacted-throughput, IPv6to4 tunnel, IPv6in4 tunnel & Dual-Stack mechanism.
1 Introduction Recent studies [1], [2] and [3] indicate that Video over IP is an important technology, which is growing rapidly and has a vital role ahead. Futuristic studies also specify that reliability and availability of Video over IP on all types of electronic devices will be on demand. Hence Video over IP would require more IP addresses in order to permit larger numbers of devices to be connected over the internet. Several other concerns are expected to arise and Video over IP has to deal with related issues, in order to enhance the performance of Video over IP. Issues like video packet size for mobile devices and quality over next generation networks (NGN) are yet to be resolved. Currently Video over IP is mainly being transmitted over IPv4 networks (Internet). However, according to researchers a greater challenge exists for transmitting video
over IP over IPv6 infrastructure. In this scenario we have implemented an infrastructure based on NGN including IPv6 to identify the quality of Video over IP using IP transition mechanisms [4]. MPEG (Moving Picture Experts Group) is a working group of specialists formed by international organisations with a view to set standards for audio, video and multimedia (MM) communications. MPEG has collaborative organisations and works with a range of universities, industries, and research institutions. MPEG standard characterizes multiple ways to broadcast audio and video such as multimedia streams that are compressed and transmitted concurrently within the MPEG standards. MPEG-1, MPEG-2 and MPEG-4 are commonly used standards form the range of MPEG standards, which are used for audio and video transmission. MPEG-3 was designed for High Definition TV compression and became redundant with its features merged with MPEG-2. MPEG-1 was the first MM compression method, which had a speed at approximately 1.5 Mega bits per second (ISO/IEC 11172). Considering the low bit rate of 1.5Mbps for MM services, this standard provides lower sampling rate for the images and uses lower picture rates of 24-30 Hz. The final outcome results a lower picture quality. The popular format known as MP3 is formed from the parts of MPEG-1 and MPEG-2 standards. MPEG-2 provides broadcast quality video and is specially used for TV transportation. The typical broadcast rates for MPEG-2 standard are higher than MPEG-1 while MPEG-4 standard uses compression techniques that result in higher throughput that is greater than MPEG-2. This aids transport of application level MM like computer graphics, animation and regular video files. In some cases MPEG-4 decoder is capable of describing three dimensional pictures and surfaces for files with .MP4 file extension. Matroska Multimedia Container, MKV is an open standard free container file format that can hold an unlimited number of video, audio, picture or subtitle tracks inside a single file. Unlike other similar formats, such as MP4, AVI and ASF, MKV has an open specification (open standard) and most of its code is open source. The formats are .MKA for audio only, .MKS for subtitles only, .MKV for audio, video, pictures and subtitles and .MK3D for stereoscopic/3D video. Matroska is also the basis for .webm (WebM) files. Matroska is based on a binary derivative of XML, called the Extensible Binary Meta Language (EBML) which bestows future format extensibility, without breaking file support in old parsers. Flash Video is viewable on most operating systems, using the Adobe Flash Player and web browser plug-ins and is very popular for embedded video on the web and used by YouTube, Google Video, metacafe, Reuters.com, and many other news providers. FLV is a container file format used to deliver video over the Internet using Adobe Flash Player (versions 6 to10). There are two different video file formats known as Flash Video: FLV and F4V. FLV was originally developed by Macromedia. The audio and video data within FLV files are encoded in the same way as they are within SWF files Flash Video content may also be embedded within SWF files. The F4V file format is based on the ISO base media file format. Flash Video FLV files usually contain material encoded with CODECS following the Sorenson Spark or VP6 video compression formats. The most recent public releases of Flash Player (collaboration between Adobe Systems and MainConcept) also support H.264 video and HE-AAC audio.
The contribution and motivation of this paper is to identify the actual-throughput and impacted-throughput using five different video protocols and compare their results. This experiment was conducted in a computer lab based on a real network. A two way video conference was established on IPv6 based networks, which were connected via IPv4 cloud. To connect these networks two tunnelling and a Dual-Stack mechanism was established between both IPv6 networks. Video traffic was transmitted using MPEG-1, MPEG-2, MPEG-4, MKV and FLV protocols over these mechanisms and actual-throughput and impacted-throughput was monitored. As of early-2011, no literature was observed that covered evaluation of video protocols on these three well known transition mechanisms such as IPv6to4, IPv6in4 and DualStack.
2 Background To resolve the issue of shortage in IPv4 addresses for the future, IPv6 was introduced to the computer world. In addition it also provides a number of other advantages on adoption. However, IPv6 still has one major issue since it does not communicate directly with IPv4 networks. To resolve this issue, researchers have designed various IP transition mechanisms known as IPv6 over 4, NAP-PT, Dual-Stack, IPv6to4 and IPv6-in-4 mechanisms, which allows IPv6 based networks to communicate with other IPv6 based networks via the IPv4 cloud. IPv6-to-4 and IPv6-in-4 are two vital tunnelling mechanisms which are available on multiple operating systems including Windows and Linux OSs. The main purpose of these tunnelling mechanisms was to enable IPv6 based networks to communicate to other IPv6 based networks through IPv4 networks (internet). The function of tunnelling mechanisms was designed to carry IPv6 packets via IPv4 networks using encapsulation process and add IPv6 packets into IPv4 header. It then executes decapsulation process at the other end and removes IPv4 header and deliver pure IPv6 based packets to their destinations. IPv6-to-4 tunnel operates as an automatic tunnel using prefixed IP addresses. A special method is used to calculate prefixed IP addresses for both IPv4 and IPv6. It also does not work with private IPv4 addresses and it cannot use multicast addresses or the loop-back address as the embedded IPv4 address [5]. IPv6-in-4 tunnel is also known as configured tunnel, which needs to be configured manually among hosts. It has the capability to operate at any given IP address and does not require any prefixed IP addresses, unlike IPv6to4 tunnel. Each of these tunnels has a special virtual interface, which requires different setup configuration. IPv6to4 tunnel is created and setup in an interface called tun6to4 whereas IPv6in4 tunnel is created and setup in an interface called IPv6-in-4. Dual-Stack mechanism is established by enabling both versions of IP (IPv4 & IPv6) protocol concurrently and they both operates simultaneously. It allows IPv4 based nodes particularly to communicate with only IPv4 based nodes while IPv6 based nodes specifically communicate with IPv6 based nodes; however IPv6 nodes can’t communicate IPv4 nodes. IP transition mechanisms have proposed a solution by allowing IPv6 based networks to communicate with other IPv6 based networks through IPv4
infrastructures. However, there are still major concerns that are noticeable with use of IP transition mechanisms. Dual-Stack is limited in application until internet service providers and other networks on the internet enable dual-stack mechanism. The tunnelling mechanisms causing additional delay in video transmission is because of encapsulation and de-capsulation process that is a vital concern. This may impact the quality of video transmission and may also lead to increased bandwidth wastage. Thus the authors have setup a real network based environment to conduct tests to clarify the impact of dual-stack and tunnelling mechanisms on five different video protocols. The IP mechanisms selected include a dual-stack, IPv6to4 and IPv6in4 tunnelling mechanisms. The video protocols investigated and tested were MPEG-1 MPEG-2, MPEG-4, MKV & FLV over Linux Ubuntu 10.10 platform. The actual aim of this experimental research is to identify the actual-throughput and clarify the impactedthroughput caused by these IP transition mechanisms on each video protocol and compare their results. The organization of this paper is as follows: Section III describes related works. Section IV presents the network setup and Section V discusses the traffic generating & monitoring tools used for this study. Section VI covers experiment design and Section VII outlines the analysis and results. Last section presents the discussions and conclusions followed by the references.
3 Related Works This section covers related areas of research which was undertaken by other researchers in past years. In [6] a method was designed and tested which purposed a solution to packet loss issue in video transmission. The method used is called Adaptive Significance Determination Mechanism in Temporal and Spatial domains (ASDM-TS) for H.264 videos packets using IP dual-stack infrastructure with DiffServ model. The video packet loss issue was undertaken in depth and multiple video protocols were involved as each protocol is based on different characteristics and experiences different errors during transmission. A model which used fixed packets for video traffic and prioritised video packet progression differently is ineffective and reduces the quality of video packets due to significant packet loss in the process of transmission. However, using this new method (ASDM-TS) can improve the packet loss in video transmission especially when it is broadcast over IP dual-stack mechanism. In this scenario different types of traffic including video was tested and analyzed on dual-stack mechanism. In [7], authors conducted an experiment and performed Video, Internet, FTP & VoIP traffic over dual-stack mechanism. The tool known as NS-2 (Network Simulator 2) was selected to carry out the tests and metrics considered were packet loss, bandwidth and delay. Video protocol involved was MPEG-4 and it was transmitted over Dual-Stack mechanism using various packet sizes and outcome was compared. It was concluded at the end, that usage of IPv6 is much better than IPv4 no matter which traffic is transmitted. Furthermore IPv6 has the capacity to transmit more bandwidth, and cause less delay for large packet sizes whereas IPv4 is limited and provides less bandwidth for large packet sizes.
Communication between two countries was setup using IPv6 to identify the behaviour of video traffic over a live network. In [8] authors observed video transmission over pure IPv6 and results obtained were compared with IPv4 based networks. The tests include HD (High Definition) video packets with and without compression system on both networks (IPv4 & IPv6) and one-way and two-way communication system was established between both countries. The traffic analysis outlines that 0.1% packet loss was measured over one-way transmission on IPv6 based networks while two-way transmission added significant packet loss at approximately 44%. The video transmission over IPv4 states that there is no major concern while using one-way and two-way video communication and outcome is stable for both. However, results for IPv6 indicates that using two-way transmission has caused significant impact on packet loss (44%) due to the network devices. Overall it was concluded that devices used in the infrastructure of IPv6 have caused this major packet loss as these device are not compatible with each other in regards to IPv6 traffic forwarding. An investigation over packet loss was conducted using video traffic. In [9] investigation was conducted to identify and compare packet loss occurrence in video transmission due to the process of error concealment and without error concealment. Lotus multi-view sequence was established that enables 8 views at a time and each view provides 500 frames. Outcome over packet loss shows that there was packet loss at approximately 2% without using error concealment process and caused significant damage to video quality. However using error concealment produced much better results and the quality of video over IP infrastructure was efficient. A new structure of carrying 3D traffic over IP networks was designed and a solution was proposed for 3D IP-TV. In [10] authors designed a technique called IP3DTV Network Management System which was established on both versions of IP (IPv4 & IPv6). Another study was carried out to enhance the performance of video over IP networks using two techniques known as SBF-H and RBF-H. The techniques mentioned above have the capability to select the appropriate packets during video transmission and forward them in bi-directional multiple lanes. The outcome was obtained based on simulated test environment. It outlines that having RBF-H technique could enhance video traffic while SBF-H is appropriate in most conditions [11]. In this paper [12] the researchers setup a network for simulation environment and performed voice and video packets over WLAN (Wireless Local Area Network) using multiple protocols. The outcome obtained from the tests shows that three different types of channels can be broadcasted concurrently without having significant packet loss in video transmission. The authors concluded at the end that the outcome achieved from these tests which was conducted in LAN (Local Area Network) environment, can be applied over WAN (Wide Area Network) without causing any impact on video quality. In [13], another study was undertaken and real-time network was established to observe the original packet loss on a live network. Impact of frame rate on real-time transmission was also investigated in [14] and [15], the research in [16] takes it to the next level by testing effects of video on next generation network (NGN) and future architectures.
4 Network Setup The proposed network test-bed was established using four different setups. First setup was based on pure IPv6, second enabled dual-stack mechanism. Third and fourth setup involved the two tunnelling mechanism known as IPv6to4 and IPv6in4. There are three networks in each setup as illustrated in Figure 1 below. Two networks at both ends are based on IPv6 configurations while the cloud is based on IPv4 configuration. To establish a connection between two IPv6 based networks through IPv4 cloud, two tunnelling and dual-stack mechanisms were configured. The two tunnelling mechanisms included are IPv6to4 and IPv6in4. One by one each of these mechanisms was configured to setup a connection between IPv6 based networks. Throughout these networks Cat5e cables were used for physical connectivity. As visible below a client workstation is connected to a router using IPv6 configuration and then a router is connected to another router using IPv4 configuration. Second router is connected to a client using IPv6 configuration. IPv6to4 and IPv6in4 tunnelling mechanisms were configured on both router machines. For Dual-Stack mechanism all the workstations and routers had both versions of IPs enabled (IPv4 and IPv6 concurrently). Linux (Ubuntu 10.10) operating system was installed on both routers and static routing was used for both versions of IPs (IPv4 & IPv6). IPv4 Network Tunnel IPv6 – IPv4 Router 1
IPv6 Sender Pc
Router 2
IPv6 Receiver Pc
Fig. 1. Network test-bed based on Tunnelling and Dual-Stack mechanisms. In addition pure IPv6 based networks were set up and similar tests performed on these networks in order to compare the results. The test-bed shown above is based on two IPv6 networks through IPv4 cloud and both IPv6 networks are connected to each other using IP transition mechanisms (IPv6to4, IPv6in4 & Dual-Stack). All tests were conducted under same circumstances using same services on each workstation. The hardware used in this experiment contains four workstations; two machines performed as clients and other two were as routers. Linux (Ubuntu 10.10) platform was installed on both router machines and three IP mechanisms were established on each of the two routers. Authors used a tool called CPU-Z to identify all the components are identical. List of hardware components is mentioned below: An Intel® Core 2 Duo E6300 1.86 GHz processor 4.00 GB RAM for the efficient operation Broadcom NetXtreme Gigabit NIC cards A Western Digital hard-drive (160 GB) on each workstation. Cat5e fast Ethernet cables were also used.
5 Traffic Generating & Monitoring Tools VLC (Video LAN Client) [17] is a tool that was selected to generate video packets over the networks. This tool was selected as it supports both versions of internet protocols (IPv4 & IPv6) and works across a range of operating systems including Linux, Windows and Mac. It also has the ability to transmit live audio, video and supports multiple voice and video protocols such as MPEG-1, MPEG-2, MPEG-4, MKV and FLV. Gnome is a network monitoring tool [18] that allows users to audit, capture and measure the status of a live network. It has the ability to detain and evaluate throughput, CPU utilization and RAM utilization. Gnome was particularly selected as it could capture and audit the video traffic during the process of encapsulation and decapsulation. Other tools tested had capability to evaluate the traffic status over a network; however they could not observe the performance of network during the process of encapsulation and de-capsulation which is operating due to IP transition mechanisms. Gnome is powerful enough and has special capacity to monitor the traffic when it is being encapsulated or de-capsulated. This tool enabled us to observe and capture the actual-throughput and impacted-throughput caused by IP transition mechanisms.
6 Experimental Design Two applications of VLC player were installed on each client workstation at both sides of networks and Gnome was setup on a router machine. First VLC application was setup to stream live video conference using one of the video protocols and it was received at other end of the network using another VLC application. Same way another VLC application was setup to stream live video back to the client, to make it two ways video conferences. Then Gnome tool was configured on a router machine where encapsulation and de-capsulation was processed. In this experiment data was captured at 30 seconds intervals. The tests were repeated over 10 times to gain more accuracy in results. Next section presents tests results obtained from this experiment.
7 Results Analysis The parameters covered in this experiment are actual-throughput, impactedthroughput and CPU utilization. This section presents performance of five different video protocols namely, MPEG-1, MPEG-2, MPEG-4, MKV and FLV over two tunnelling and Dual-Stack mechanism and their average results are shown in graphs and Table 1 below. Actual-throughput: This is original throughput of two way video conference with no additional traffic impact due to encapsulation. It simply carries video packets and delivers them to their destinations with no addition to the packet size.
Impacted-throughput: This is caused by IP transition mechanisms and is added into actual-throughput. The addition of additional encapsulation in the network wastes significant amount of bandwidth to deliver actual-throughput. Figure 2 below illustrates the actual-throughput on IPv6 and impacted-throughput due to encapsulation process on IPv6to4 tunnelling mechanism. The results were obtained using five different video protocols. Highest impacted-throughput was identified on MPEG-2 at approximately 242 kilobytes per second additional impact on actual-throughput. Second highest was calculated on MKV at approximately 136 kilobytes per second. Lowest amount of impacted-throughput was observed on FLV at approximately 17 kilobytes per second while second lowest impacted-throughput was measured on MPEG-4 at approximately 51 kilobytes per second. MPEG-1 had actual-throughput of approximately 257 kilobytes per second and additional 106 kilobytes per second more due encapsulation. MKV and MPEG-1 had almost the same amount of actual-throughput kilobytes per second; however impactedthroughput shows that MKV was more impacted than MEPG-1 at approximately 30 more kilobytes per second. Impacted-throughput is actually wasting bandwidth in a real network as it uses more bandwidth to send less amount of throughput, which is costly for users.
Throughput (KiB/s)
Throughput vs Impact-Throughput
IPv6to4 IPv6
750 600 450 300 150 0
MPEG-1
MPEG-2
MPEG-4
MKV
FLV
Video Protocols
Fig. 2. Actual-Throughput with IPv6 and Impacted-Throughput by IPv6to4 tunnelling mechanism. (KiB/s) The results analysis using IPv6in4 tunnelling mechanism indicates that it has significant impact on the protocols tested. The measurement of MPEG-2 reveals that it was impacted most as it has larger packet size and impacted-throughput was observed at approximately 246 Kilobytes per second. Least amount of impactedthroughput was calculated on FLV at approximately 17 kilobytes per second while second lowest was measured on MPEG-4 at approximately 70 kilobytes per second. MPEG-1 had actual-throughput for two way video conference at approximately 257 kilobytes per second while 110 kilobytes per second impacted-throughput was added on it. The results for MKV shows that it almost had similar amount of actualthroughput as MPEG-1 while impacted-throughput measurement shows it was impacted even more than MPEG-1 at approximately 16 more kilobytes per second.
Throughput (KiB/s)
Throughput vs Impact-Throughput
IPv6in4 IPv6
750 600 450 300 150 0
MPEG-1
MPEG-2
MPEG-4
MKV
FLV
Video Protocols
Fig. 3. Actual-Throughput with IPv6 and Impacted-Throughput by IPv6in4 tunnelling mechanism. (KiB/s) The results for Dual-Stack mechanism are presented below in Figure 4. It illustrates that dual-stack has very less impact over actual-throughput using all five protocols. Highest impact was measured using MKV protocol as it produced 15 kilobytes per second impacted-throughput while second highest was calculated on MPEG-1 at approximately 4 kilobytes per second. Lowest amount of impactedthroughput was observed on FLV at approximately 0.9 kilobytes per second. MPEG-2 and MPEG-4 both provided marginally close impacted-throughput having less than 0.2 kilobytes per second difference. The reason dual-stack has less impact on these protocols, is because it does not process encapsulation and de-capsulation.
Throughput (KiB/s)
Throughput vs Impact-Throughput
Dual-Stack IPv6
500 400 300 200 100 0
MPEG-1
MPEG-2
MPEG-4
MKV
FLV
Video Protocols
Fig. 4. Actual-Throughput with IPv6 and Impacted-Throughput by Dual-Stack mechanism. (KiB/s) Table 1: CPU Utilization tested over Tunnelling and Dual-Stack mechanisms. (%) Mechanisms Dual-Stack
CPU Utilization% MPEG-1
MPEG-2
MPEG-4
MKV
FLV
30.0
28.1
33.4
29.1
26.3
26.6 27.1 33.3 30.1 26.2 27.4 25.1 31.7 28.9 26.4 Table 1 above shows results of CPU utilization. These results were obtained during the performance test of each video protocol using the two tunnelling and DualStack mechanism. The results for IPv6to4 and IPv6in4 didn’t show much inconsistency for CPU usage while Dual-Stack was slightly higher than both of them. IPv6to4 IPv6in4
It is because both versions of IP operate concurrently. Comparison between protocols also didn’t have much variation. However, results for MPEG-4 were marginally higher than all four protocols (MPEG-1, MPEG-2, MKV & FLV). It is due to high compression method used to process MPEG-4 protocol. Least amount of CPU was used during FLV tests, as it can be seen from Table 1 above that FLV had 26 percent usage of CPU no matter which mechanisms it was tested on. It is due to the size of this protocol which requires less processing power.
8 Discussion & Conclusion In this paper investigation of actual-throughput and impacted-throughput was identified and results compiled were presented in the section above. The results clarified that video packets with larger size have significant impact while video protocols with smaller size had slight impact. Dual-Stack had had only very slight impact on all the protocols tested as it does not process encapsulation and decapsulation. On the contrary the tunnelling mechanisms had significant impact on all the protocols tested. Performance of dual-stack was much better; however it will take several years before websites and other internet service providers allow dual-stack mechanism to be used over the internet. In recent years IPv6to4 and IPv6in4 tunnelling mechanism will enable IPv6 based networks to communicate; however the impact caused by these two tunnelling mechanisms is significant and wastes a lot of bandwidth for video transmission. IPv6to4 mechanism marginally performed better than IPv6in4 mechanism on all the video protocols tested. Comparison between video protocols indicates that FLV protocol was the least impacted, and usage of this protocol will not cause much bandwidth wastage while MPEG-4 was second best for use over tunnelling mechanisms. CPU utilization measurement shows no additional impact on CPU usage while IP transition mechanisms were used. However MPEG-4 was marginally higher than other four video protocols tested on two tunnelling and Dual-Stack mechanisms. Due to the packet size of FLV it needs less CPU power. So it is clear from Table 1 above that IP transition mechanisms had slight impact on CPU processing power and even encapsulation and de-capsulation process caused insignificant impact on CPU utilization. Future work in this area should also include study and comparison of alternative methods that could be used to forward IPv6 traffic on IPv4 core networks. Another area is to cover other metrics such as jitter and packet loss and increase traffic loads of these protocols to relate the experiments to the realistic environments.
9 Acknowledgments We would like to show appreciation to UNITEC, Institute of Technology for supporting the research team and providing us this opportunity to fulfil this research.
References 1. Atenas, M., Garcia, M., Canovas, A and Lloret, J.: MPEG-2/MPEG-4 Quantizer to Improve the Video Quality in IPTV Services. IEEE Sixth International Conference on Networking and Services. pp 49-54, (2010) 2. Schierl, T., Gruneberg, K and Wiegand, T.: Scalable Video Coding Over RTP and MPEG-2 Transport Stream in Broadcast and IPTV Channels. IEEE Journals on Wireless Communications. 16 (5), pp. 64-71, (2009) 3. Kim, S and Yongik, Y.: Video Customization System using Mpeg Standards. IEEE International Conference on Multimedia and Ubiquitous Engineering. pp. 475-480, (2008) 4. Tao, S., Apostolopoulos, J & Guerin, R.: Real-Time Monitoring of Video Quality in IP Networks. IEEE/ACM Transactions on Networking. 16(5), pp. 1052, (2008) 5. Stockebrand, B.: IPv6 in Practice. A Unixer’s Guide to the Next Generation Internet. Springer Berlin Heidelberg: New York, (2007) 6. Lee, C., Yu, Y and Chang, P.: Adaptable Packet Significance Determination Mechanism for H.264 Videos over IP Dual Stack Networks. IEEE 4th International Conference on Communications and Networking. pp. 1-5, (2009) 7. Sanguankotchakorn, T and Somrobru, M.: Performance Evaluation of IPv6/IPv4 Deployment over Dedicated Data Links. IEEE Conference on Information, Communications and Signal Processing. pp. 244-248, (2005) 8. Lee, L and Chon, K.: Compressed High Definition Television (HDTV) over IPv6. IEEE Conference on Applications and the Internet Workshops. PP. 25, (2006) 9. Zhou, Y., Hou, C., Jin, Z., Yang, L., Yang, J & Guo, J.: REAL-TIME Transmission of High-Resolution Multi-View Stereo Video over IP Networks. IEEE Conference: The True Vision-Capture, Transmission and Display of 3D Video. pp. 1, (2009) 10.Luo, Y., Jin, Z and Zhao, X.: The Network Management System of IP-3DTV Based on IPV4/IPV6. IEEE 6th Conference on Wireless Communications Networking and Mobile Computing. pp. 1-4, (2010) 11.Xie, F., Hua, K.A., Wang, W and Ho, Y.H.: Performance Study of Live Video Streaming over Highway Vehicular Ad hoc Networks. IEEE 66th Conference on Vehicular Technology. pp. 2121-2125, (2007) 12.Gidlund, M and Ekling, J.: VoIP and IPTV Distribution over Wireless Mesh Networks in Indoor Environment. IEEE Transactions on Consumer Electronics. 54(4), 1665 – 1671. (2008) 13.Kukhmay, Y., Glasman, K., Peregudov, A and Logunov, A.: Video over IP Networks: Subjective Assessment of Packet Loss. Tenth IEEE Conference on Consumer Electronics. pp. 1-6, (2006) 14.Khalifa, N.E.-D.M. and Elmahdy, H.N.: The Impact of Frame Rate on Securing Real Time Transmission of Video over IP Networks. IEEE Conference on Networking and Media Convergence. pp. 57-63, (2009) 15.Sims, P.J.: A Study on Video over IP and the Effects on FTTx Architectures. IEEE Conference on Globecom Workshops. pp. 1-4, (2007) 16.IlKwon, C., Okamura, K., MyungWon, S and YeongRo, L.: Analysis of Subscribers’ Usages and Attitudes for Video IP Telephony Services over NGN. 11th IEEE Conference on Advanced Communication Technology. pp. 1549-1553, (2009) 17.Video LAN.: Video LAN Organization: VLC Media Player. http://www.videolan.org/vlc/, (2011) 18.GNOME.: GNOME Documentation Library: System Monitor Manual. http://library.gnome.org/users/gnome-system-monitor/, (2011)
